| Summary: | 碩士 === 國立成功大學 === 資訊管理研究所 === 105 === Tourism has become an indispensable part of life. There are many online travel related blog articles. Due to massive information, online readers tend to watch their favorite bloggers articles. Popular bloggers are called opinion leaders. Compared to other commentators, opinion leader has high interaction. Their influence could not be ignored.
Popular bloggers are often ranked by popularity, and it does not fully reflect the usefulness of the article. This study analyzes the sentiment and trend of the articles of the bloggers and establishes the influence network by the subscribing relationship. Combining the features above, the opinion leaders ranking will be more in line with expectations. In the field of tourism, adopting location to be topic can be more close to the needs of users. Through the use of blogger defined tag to analyze their expertise, the ranking results can be presented by topic.
Through this study, users can choose target bloggers based on opinion leader ranking results which presented by topic. They could browse these bloggers’ articles to make decision more efficient. The experimental results show that using social influence score to identify opinion leader could get the best result compared with other scores. Finally, this study identified opinion leader by combining all scores in to one. The ranking results could perform better than using social influence score.
|
|---|
